Dear users and developers,
We are trying to model a polycrystalline sample consists of few (25%) Ni3Zr grain (has HCP lattice structure) and rest of the grains (75%) are pure Ni (has FCC lattice), with random orientation across the boundary of Ni3Zr and Ni. We have prepared our desired model by using atomsk code and have exported to lammps data format.
We are using Mendelev EAM potential (Philosophical Magazine 95, 224-241 (2015)) for this system.
-------------------------Lammps set up ----------------------------
pair_style eam/fs
pair_coeff * * parameters.dat Ni Zr
velocity all create 300 72489 dist gaussian mom yes rot no
fix 1 all nvt temp 300 300 0.1 drag 1.0
Dear users and developers,
We are trying to model a polycrystalline sample consists of few (25%)
Ni3Zr grain (has HCP lattice structure) and rest of the grains (75%) are
pure Ni (has FCC lattice), with random orientation across the boundary of
Ni3Zr and Ni. We have prepared our desired model by using atomsk code and
have exported to lammps data format.
We are using Mendelev EAM potential (*Philosophical Magazine 95, 224-241
(2015)*) for this system.
-------------------------Lammps set up ----------------------------
pair_style eam/fs
pair_coeff * * parameters.dat Ni Zr
velocity all create 300 72489 dist gaussian mom yes rot no
fix 1 all nvt temp 300 300 0.1 drag 1.0
------------------------------------------------------------
-------------------
But, during the initial equilibration run, after few hundreds of steps the
entire model being thermalized nonphysically (T_initial~300 K and
T_final~5000 K) and leads to amorphization along with small voids
nucleation in the boundary.
Now our question is, is this EAM potential is capable to describe Ni3Zr
and as well as pure Ni phase concurrently ? If not, then what would be
possible way out ?
impossible to say from remote. how do you know, that you didn't make an
error when constructing the potential file? how do you know that your
initial geometry and your MD settings are proper and suitable for your
system?
before modeling your specific system, have you run any tests to validate
the potential by reproducing calculations given in the reference where you
took the parameters from?
axel.
Dear Prof. Kohlmeyer,
We did some initial calculation with this same potential taking pure Ni3Zr phase and the mechanical properties obtained from this EAM is in good agreement with experimental reported values. But, problem occurs when we are trying to mix (making polycrystals) two different phases, like, pure Ni3Zr and pure Ni. Therefore, our question is whether a EAM potential (fitted for Ni and Zr alloys) would be capable to describe Ni3Zr(pure)+Ni(pure) system ?
Thanks for your time.
Best,
Rajdeep
Dear Prof. Kohlmeyer,
We did some initial calculation with this same potential taking pure Ni3Zr
phase and the mechanical properties obtained from this EAM is in good
agreement with experimental reported values. But, problem occurs when we
are trying to mix (making polycrystals) two different phases, like, pure
Ni3Zr and pure Ni. Therefore, our question is whether a EAM potential
(fitted for Ni and Zr alloys) would be capable to describe
Ni3Zr(pure)+Ni(pure) system ?
well, that is something you can best answer yourself by making more tests,
e.g. for pure Ni. and simpler systems.
there is not enough information (actually, none at all) here to
convincingly prove that the potential might be the cause for your
observation an not, e.g. a bad geometry with overlaps or bad simulation
parameters or one of the many other reasons that can make MD simulations do
unexpected things.
besides, your question is about the physics about EAM potentials and not at
all about LAMMPS, so it is rather off-topic.
axel.